cerebras.modelzoo.data.nlp.bert.BertClassifierDataProcessor.DataProcessor

class cerebras.modelzoo.data.nlp.bert.BertClassifierDataProcessor.DataProcessor

Bases: abc.ABC

Base class for processors that load their raw data from TFDS. Child classes must provide map_fn, name.

Parameters

• data_params (dict) –

  Input parameters for creating dataset. Expects the following fields:

  • ”vocab_file” (str): Path to the vocab file.

  • ”data_dir” (str): Path to directory containing the TF Records.

  • ”batch_size” (int): Batch size.

  • ”max_sequence_length” (int): Maximum length of the sequence.

  • ”shuffle” (bool): Flag to enable data shuffling.

  • ”shuffle_seed” (int): Shuffle seed.

  • ”shuffle_buffer” (int): Shuffle buffer size.

  • ”do_lower” (bool): Flag to lower case the texts.

  • ”num_workers” (int): How many subprocesses to use for data loading.

  • ”drop_last” (bool): If True and the dataset size is not divisible

    by the batch size, the last incomplete batch will be dropped.

  • ”prefetch_factor” (int): Number of samples loaded in advance by each worker.

  • ”persistent_workers” (bool): If True, the data loader will not shutdown

    the worker processes after a dataset has been consumed once.

• model_params (dict, optional) – Model parameters for creating the dataset, unused.

Methods

create_dataloader
create_dataset

__init__(data_params, model_params) → None

previous

cerebras.modelzoo.data.nlp.bert.BertClassifierDataProcessor.ClassifierDataset

next

cerebras.modelzoo.data.nlp.bert.BertClassifierDataProcessor.MNLIDataProcessor